Learning Disabilities as Operationally Defined by Schools
Donald L. MacMillan, University of California, Riverside, & Gary N. Siperstein, University of Massachusetts, Boston
Learning Disabilities Summit: Building a Foundation for the Future White Papers
This paper is available in alternative formats: | Download Word | Download pdf |
Pages: | 1 | 2 | 3 | 4 | 5 | 6 | 7 |
TITRATION OF INTENSITY OF TREATMENTS IN DECISION MAKING
Currently, the LD category appears to be used by the schools as a general, nonspecific category embracing students who can best be characterized as exhibiting absolute low achievement. In some ways, its use by the schools negates the need for the "developmental delay" category permitted for younger students under IDEA guidelines. Elsewhere we (MacMillan, Gresham, Bocian, & Siperstein, 1997) have described the decision-making process in the public schools as one that uses resistance to treatments as its guiding principle. Figure 2 is a schematic representing how treatments are titrated and how a child resistant to one treatment is subsequently moved to a treatment that is more intense until the child ultimately receives effective instruction or is reclassified. We characterized the application of increasingly more "intensive" treatments as a titration process. Resources available in the general education classroom are rather weak treatments and a child failing to make adequate progress might receive Chapter I services. However, if inadequate progress persists, the child is referred for prereferral interventions, individually tailored interventions are implemented, and the impact is evaluated. If these prereferral interventions are judged to be ineffective, the child is typically referred by a team for formal evaluation and assessment (described in detail elsewhere in this paper).
The initial diagnosis is commonly LD, often despite a failure to exhibit criteria specified, and the IEP calls for an initial treatment of pullout resource service for a limited duration (e.g., 2 to 4 hours per week). If this action fails to remediate the academic weaknesses, the duration of resource help is increased (e.g., to 6 to 8 hours per week) or a child is placed into a special day class--in some way providing for more intensive academic and remedial treatments. For a segment of the SI LD population receiving the most intensive LD services available, a failure to respond is still noted. In Figure 2 these cases are sometimes reclassified as MR or ED. Recall the experiences of ED students described by Duncan et al. (1995) who were initially classified as LD only to be reclassified as ED later in their school careers.
In the titration model described herein, the data coming from formal assessments contribute little to the decision-making process; instead, they are necessarily collected in order to comply with regulations. IQ data are ignored when they point to mental retardation as a diagnosis and also when establishing "expected" levels of reading or arithmetic achievement for purposes of LD eligibility. This raises the issue of how, if at all, intelligence testing should be used in the process.
ISSUE OF INTELLIGENCE TESTING
Few things have been as hotly debated as the role of IQ, and we do not intend to rehash that history here. Instead, we urge some focused discussion of the usefulness of IQ in classification and informing intervention efforts. Our reading of the literature on discrepant versus nondiscrepant low achievement (see Fletcher et al., 1998) in students with IQ scores of 80 or above seems to conclude that intelligence tests have limited usefulness for the identification of students with LD. Yet administration of such tests is currently mandated for establishing eligibility in most states for students as LD or MR--at what cost? Gresham and Witt (1997) wrote:
Estimates suggest that between 1 and 1.8 million intelligence tests are administered individually to children each year in the United States. Recent survey data suggest that two-thirds of a school psychologist's time is spent in special education eligibility determination and the typical school psychologist administers over 100 individual tests of intelligence each year. (p. 249)
Evidence presented throughout this paper points to the fact that the IQ data are not used in making differential diagnoses. Moreover, even the most ardent defendant of intelligence testing would probably concede that given the omnibus nature of the test, it has no "curricular relevance" (i.e., does not inform us as to what instructional strategies will work). If IQ scores are not used in a consistent fashion for purposes of classification and they are not instructionally relevant, does it make sense to mandate their continued use on a wholesale basis merely to establish eligibility? We find little evidence to support their continued use on a wholesale basis.
Figure 2. Titration of intensity of treatments.
The one issue that must be considered, however, if the routine administration of IQ tests is discontinued, is the importance of exclusionary criteria. That is, if a reconceptualization of LD continues to exclude conditions like mental retardation, then the grounds on which to make the exclusionary decision warrants consideration. If a defining characteristic of mental retardation is "low general intelligence," then there will have to be some basis on which to make that determination. Abundant evidence has been presented (e.g., MacMillan, Gresham, Siperstein, & Bocian, 1996) showing that even when IQ data are presented to document mental retardation and permit the exclusion, the schools ignore it in most instances and do not enforce the exclusionary criteria. Another reasonable position is to adopt a domain-specific approach to what is now LD (e.g., reading disability, math disability) and, without establishing the presence of mental retardation, use an approach consistent with that proposed by Lynn Fuchs in which students with a reading disability, for example, are exposed to best-practice treatments implemented with integrity. For those cases who fail to respond favorably to a validated intervention implemented with integrity, consider the consequences of persisting in pursuing achievement in reading (and side effects of continued failure) and consider providing that child with a more functional curriculum. Does it matter for educational decision making whether a nonresponder has a low IQ or such a severe problem independent of low IQ?
The ultimate decision regarding the use of IQ in classification for school purposes ought to be made after careful consideration of the consequences of the alternatives.
EFFORTS TO "FIX" LD: THE NEED FOR A BROADER PERSPECTIVE
We have seen an ongoing debate over the definition of LD for more than a quarter of a century. During this same time, special education has become more "specialized" in the sense that the generalists have become fewer and fewer and those in our field identify themselves in terms of subspecialties within a disability category (e.g., my area is memory within the LD field). However, as we saw in 1973, a change in the definition of mental retardation was undertaken without consideration of the consequences of this change on the LD population. From the perspective of the schools, the "judgmental categories" (including LD, MMR, ED, speech and language impaired; other health impaired) are being used in idiosyncratic ways in order to serve children the schools believe need special education services. Any attempt to "fix" the LD definition and criteria that fails to consider the criteria for other judgmental categories and issues of comorbidity is, in our opinion, doomed. Furthermore, it is crucial to acknowledge the differences between urban and suburban schools and the implications of these differences for the educational process.
Students served currently as LD, MMR, and ED overlap considerably along certain behavioral dimensions. In terms of reading, students served in all three groups tend to exhibit reading disabilities. Moreover, within a given disability group, the degree of reading disability varies considerably. Another behavioral dimension that is salient to all three categories is externalizing behavior problems. It is a defining characteristic for many ED students, but the frustration experienced by LD students appears to give rise to externalizing behavior problems for many LD and ADHD students (see review by Hinshaw, 1992). Viewed as a Venn diagram, the overlap between members of these disability groups is considerable. A third dimension relevant to all three is the relative weakness in social skills and peer relationships and the frequency with which children in these categories experience social rejection (see Asher & Coie, 1990). As suggested above, the relevance of intelligence is debatable, but might warrant consideration. At present, a categorical approach is used in which children are placed into one, and only one, of the extant disability categories. Yet, a child categorized as LD may, in addition to problems in reading, exhibit significant externalizing behavior problems. Current criteria for ED require that the behavior problems or emotional problems must adversely affect the child's academic performance. Hence, ED students require effective treatments for behavior and academics.
Our point here is that a multidimensional approach to assessing behavioral dimensions salient to all three current disability categories might provide assessments directly relevant to the treatment program. It may further gauge the severity of a given child's problem on a given dimension that would inform those crafting the IEP about the extent to which a given behavior dimension should be addressed in programming. The current assessments measuring static variables often unrelated to treatment protocols fail to capture opportunity to learn as a competing explanation for the low achievement. To the extent that the LD category has embraced students whose low achievement appears linked to experiences of poverty, the issue of sociocultural factors as causes or contributors to the poor achievement simply cannot be ignored. We must recognize that factors of impoverished learners contribute to both learning problems in children and how the eligibility process is compromised in schools serving children of poverty.
LD AND SOCIAL CLASS
As noted previously, urban and suburban schools serve LD populations that differ distinctly from one another. The work of Gottlieb et al. (1994) on urban LD students led to the following conclusion: "Data we have collected over a 10-year period indicate that today's child with learning disabilities functions very similarly to the way students with educable mental retardation performed 25 years ago" (p. 453). This finding raises a number of questions for the special education delivery system and those interested in students with LD. Consider that the condition of mild mental retardation is almost exclusively a phenomenon of poverty. Richardson's (1981) research, conducted in Aberdeen, Scotland, where all subjects were White, provided clear evidence that the form of mental retardation with IQ scores above 50 and no evidence of CNS involvement was simply not found in the highest social class strata. Richardson plotted the prevalence of this form of mild mental retardation against the prevalence of two other forms (IQ < 50, evidence of CNS involvement; IQ > 50 and evidence of CNS involvement). In the two lowest social class strata, this form of mental retardation (IQ > 50 and no CNS involvement) constituted the single largest proportion of cases.
In this paper we have provided evidence that children resembling the Richardson cases of mild mental retardation are among the most frequently referred students (MacMillan, Gresham, Siperstein, & Bocian, 1996) and that in urban settings, children with IQ < 85 constitute more than half of the LD population in urban districts. These findings raise a couple of questions in our minds, and we do not know of any data set that provides any answers to the questions. First, in urban districts in which low cognitive children constitute the most visible form of learning problems, are students with the traditional LD profile of low achievement despite average or above-average IQ not being served by special education? We hypothesize that the low cognitive students require the most accommodations in a general education class and therefore deviate most markedly from the model student profile. Teachers in these settings refer these students but do not refer the "traditional LD" cases, or if they do, the committee charged with establishing eligibility uses the available slots in LD programs to serve the low cognitive children. As we discussed previously, this form of false negative case has not been studied and the magnitude of this group is unknown. It would be interesting to know whether fewer cases of false negative LD cases are to be found in affluent suburban districts than in urban districts, given that one would not expect to find the old EMR and borderline children in suburban districts. A second question concerns the proportion of a district's student body that one would expect in the overall special education program. If urban and suburban districts serve roughly the same proportion of their students in the nonjudgmental categories (e.g., visually impaired, orthopedically impaired), but urban districts also serve low cognitive cases associated with poverty and traditional LD students, one might hypothesize that a higher proportion of an urban district's student bodies are in need of special education services than is true for suburban districts.
The process prescribed under IDEA plays itself out in very different ways in different school districts. A failure to recognize this leads to false assumptions about the nature of LD students. Moreover, we must come to grips with the realities that school districts serve different populations of children, have differing resources to address problem learners, and make eligibility decisions in light of these different circumstances. At present, schools do not identify cases that consistently fit the idealized models described in authoritative definitions or state education codes. This situation may frustrate the research community and others removed from the front lines of education. At the same time, we know of precious little evidence suggesting that the children who are served as LD in these diverse districts are not in need of help. Any resolution of this state of affairs, in our opinion, must begin with consideration of all judgmental disability categories, not just LD.
CONCLUDING REMARKS
We urge recognition of one reality driving the public schools' focus on planning for services: They are going to continue serving those students they perceive to be the most in need of help. At present, the way they are serving those most in need is by using the LD category as the vehicle for providing the help they perceive as needed. Doing so has resulted in increasingly less and less overlap between the population of children the schools serve as LD and that described in authoritative definitions and state education codes of LD, particularly in urban schools. We must acknowledge the existence of a large segment of marginalized students, many of whom encounter learning difficulties for reasons other than intrinsic, neurologically based causes. Moreover, the public schools recognize this large undifferentiated group of students with achievement deficits, use the LD category to justify serving them, and do so on the basis of absolute low achievement, not "discrepant low achievement." Those whose professional interests reside with the traditional LD student would be well advised to acknowledge the educational needs of the nontraditional LD, join forces with those who advocate for serving these "false positive" LD children, advocate for their being served, and engage in a discussion with advocates for these nontraditional LD students in order to secure appropriate services for them while recognizing and acknowledging differing etiologies and presumably differing educational needs. A failure to do so will, in our opinion, result in a continuation of the current state of affairs, clouding the parameters of the LD category, because many of these nontraditional LD children are among the "most difficult to teach" and will be a priority among public school teachers.
There exists an unhealthy schism between research and practice fueled, in part, by the discrepancy between SI and RI students with learning disabilities. Public school personnel perceive the research community as out-of-touch while the research community often views those in the public schools as uninformed. In truth, the research does not inform practice as the data base derives from a population of "LD" students only vaguely resembling SI "LD" students. An analogy to medicine may clarify our point. Research on the treatment of diabetes informs physicians treating diabetics because the researchers and the practitioners agree on who is diabetic. Researchers studying subjects with LD and the practitioners serving students with LD do not agree on who is LD and, as a result, research does not inform practice.
It is our probably naive belief that efforts to revise definitions of judgmental disability categories should begin with "low achievement due to..." and then acknowledge that in our best clinical judgment the low achievement is apparent due to one of several factors. Among the factors currently confusing the LD category are (a) low general intelligence, (b) emotional/behavioral conditions, (c) specific processing difficulties, (d) environmental disadvantage, and/or (e) lack of opportunity to learn, particularly because of inadequate instruction. This position obviously favors increased refinement, or differentiation of, categories as opposed to "noncategorical categories." To that end we would argue that one-time assessments cannot make such distinctions as they tap static variables that are insensitive to such distinctions. Instead we would argue for multiple assessments of progress, using measures/scales sensitive to change in response to interventions implemented with integrity. In essence, progress monitoring of achievement after exposure to best-practice treatments intimately linked to the very achievement deficits prompting referral would provide the basis for eligibility decision making. Doing so would require revisions in eligibility criteria aligning the new assessments with the primary concerns of the public schools and tapping those achievement deficits targeted in the reading disability and math disability research. We believe that response to known treatments would begin to further clarify the varied etiologies of learning difficulties and create categories with greater validity (i.e., what one knows about cases falling into each of the categories specified above). In turn, cleaning up the "aptitudes" in the equation would enhance examination of aptitude treatment interactions. One thing is certain; you will never escape the "hall orf mirrors that extends to infinity" noted by Cronbach (1975, p. 119) as one studies interactions if the aptitudes are ill-defined. At present LD is ill-defined.
